AIM Step 10
Student Memory Support
Inferential Biostatistics: Confidence Intervals, Hypothesis Testing and Statistical Tests
4th Year MBBS • Neuroscience • High-yield memory reinforcement and last-minute revision
Step 10 format follows the supplied AIM Student Memory Support specification. :contentReference[oaicite:0]{index=0}
1. High-Yield Flashcards
Tap each question to reveal the answer.
What does the standard error of the mean describe?
The precision of the sample mean as an estimate of the population mean.
How does increasing sample size affect standard error?
Standard error decreases because SE = SD ÷ √n.
What is the general structure of a confidence interval?
Estimate ± critical value × standard error.
What does a narrower confidence interval indicate?
Greater precision of the estimated population parameter.
What does a P value represent?
The probability of the observed or more extreme result assuming the null hypothesis is true.
What is a Type I statistical error?
Rejecting a null hypothesis that is actually true.
What is a Type II statistical error?
Failing to reject the null hypothesis when a real difference or association exists.
How is statistical power related to beta?
Power = 1 − β.
When is a Z-test framework appropriate?
When the sampling distribution can reasonably use the normal approximation, including suitable large-sample mean or proportion testing.
Why is the t distribution used when population standard deviation is unknown?
Population variability is estimated from the sample, creating additional uncertainty.
Which t test compares repeated measurements in the same participants?
Paired t test.
What does the chi-square statistic compare?
Observed frequencies with frequencies expected under the null hypothesis.
What does a correlation coefficient near −1 indicate?
A strong negative linear relationship.
What does the slope in simple linear regression represent?
The expected change in the outcome for a one-unit increase in the predictor.
What should be checked before trusting an SPSS statistical result?
Variable type, coding, missing values, outliers, study design, distribution and test assumptions.
2. Mnemonics
Mnemonic Title: Choosing the t Test
PIO
Meaning: Paired = same subjects or linked measurements; Independent = separate groups; One-sample = one sample mean versus a reference mean.
Mnemonic Title: Chi-Square Core Calculation
OE
Meaning: Think Observed versus Expected frequencies whenever you see a chi-square question.
Mnemonic Title: Before Choosing a Statistical Test
DAD
Meaning: Data type → Assumptions/distribution → Design, especially paired versus independent observations.
3. Memory Tables
Common Test Selection
| Method | Main Data | Main Purpose | Memory Clue |
|---|---|---|---|
| Z test | Mean/proportion under suitable normal approximation | Compare estimate with null value | Standard normal distribution |
| t test | Continuous means | Compare one or two means | Population SD estimated |
| Chi-square | Categorical counts | Association or goodness of fit | Observed versus expected |
Correlation versus Regression
| Feature | Correlation | Regression |
|---|---|---|
| Main question | How strongly do variables move together? | How does outcome change with predictor? |
| Main output | Correlation coefficient | Regression equation and slope |
| Interpretation | Strength and direction | Expected change or prediction |
| Caution | Does not prove causation | Model must be interpreted within appropriate data |
4. Rapid Revision Points — Last-Minute Revision
Must Remember:
- Standard deviation describes spread among observations; standard error describes precision of an estimate.
- Higher confidence level produces a wider confidence interval when the data remain unchanged.
- A confidence interval including the null value does not show a statistically clear effect at the corresponding significance level.
- P < α supports rejection of H₀; P ≥ α means insufficient evidence to reject H₀.
- A non-significant result does not prove that two treatments or groups are identical.
- Lower α reduces Type I error risk but may increase Type II error if other study features remain unchanged.
- A right-skewed distribution has a long tail toward higher values.
- Choose parametric or non-parametric analysis only after considering data type, design and assumptions.
- Very small expected cell counts may make the chi-square approximation unreliable.
- Statistical significance and clinical significance answer different questions.
Common KMU Trap: A P value of 0.03 does not mean there is a 3% probability that the null hypothesis is true.
5. Clinical Memory Hooks
Large trial + 1 mmHg difference + very small P value
→
Statistical significance may coexist with a clinically small effect.
→
Statistical significance may coexist with a clinically small effect.
Same patients measured before and after intervention
→
Linked observations → paired t test.
→
Linked observations → paired t test.
Smoking status compared with presence or absence of chronic cough
→
Two categorical variables → chi-square test of association.
→
Two categorical variables → chi-square test of association.
Strong Pearson correlation driven by one extreme observation
→
Inspect the scatterplot because an outlier can strongly influence correlation.
→
Inspect the scatterplot because an outlier can strongly influence correlation.
6. Starred High-Yield Exam Points
- ⭐ SE = SD ÷ √n; increasing sample size improves precision.
- ⭐ 95% CI under the normal approximation = estimate ± 1.96 × SE.
- ⭐ Type I error = false-positive statistical conclusion; Type II error = false-negative statistical conclusion.
- ⭐ Paired t test analyzes the mean of within-subject differences.
- ⭐ Chi-square uses categorical frequencies and compares observed with expected counts.
- ⭐ Correlation does not establish causation; regression slope estimates expected outcome change per unit predictor change.